Kevin Rooney, BS, Jenna Adalbert, MPH Asif M. Ilyas, MD, MBA, FACS
In 2019, PA ranked 26th in state opioid prescription rates with 47 prescriptions per 100 people, just above the U.S. average of 46.7 prescriptions
Of the 16 PA counties with the highest total or adjusted per 100,000 population hydrocodone and oxycodone prescriptions, half of them had fewer PCPs and non-physician providers than average to serve their population ratio
Those with opioid use disorder who had at least one primary care visit in the past year were more likely to enroll in medication-assisted therapy (MAT) and filled more buprenorphine and naltrexone prescriptions than those without a PCP visit during the same period
Increased PCP access is an important opioid harm reduction measure for vulnerable populations through interventions such as improved patient education and counseling, opioid prescription dosage/amount monitoring and opioid dosage tapering
Opioid prescriptions reported by the Pennsylvania (PA) Prescription Drug Monitoring Program (PDMP) have continued to decrease steadily since the PA Department of Health assumed control of PDMP operations in 2016. In 2019, PA ranked 26th in state opioid prescription rate with 47 prescriptions per 100 people, just above the U.S. average of 46.7 prescriptions (2). However, research has shown that there is large variation in opioid prescription rates between counties in PA (7). Accordingly, where are these discrepancies found in PA and how do primary care providers, the highest prescribers of opioids, fit into these trends?
Opioid Prescription Trends by County
When analyzing total prescriptions for both hydrocodone and oxycodone per county in PA for the first quarter of 2020, it is expected that the five highest counties are predominantly the most heavily populated areas in the state. For hydrocodone prescriptions, Allegheny, Erie, Westmoreland, Philadelphia, and Lancaster dispensed the greatest total number of prescriptions (Figure 1).
Figure 1. Top Five Counties for Hydrocodone Prescriptions by Total Volume
For total oxycodone prescriptions during the same period, Philadelphia, Allegheny, Montgomery, Bucks, and Delaware County dispensed the greatest total number of prescriptions (Figure 2) (7).
Figure 2. Top Five Counties for Oxycodone Prescriptions by Total Volume
Each of these counties is defined as an "urban” area by the 2010 U.S. Census Data (1). However, when prescription rates were adjusted per 100,000 population, a different trend was observed. Following adjustment, the top five counties for hydrocodone prescription volume consisted of Erie, Fayette, McKean, Clinton, and Blair County (Figure 3).
Figure 3. Top Five Counties for Hydrocodone Prescriptions Adjusted per 100,000 Population
For oxycodone, Fayette, Washington, Carbon, Northumberland, and Cambria were the highest prescribing counties during this period after adjusting for population (Figure 4) (7).
Figure 4. Top Five Counties for Oxycodone Prescriptions Adjusted per 100,000 Population
While urban centers initially appeared to outrank other county census classifications in total hydrocodone and oxycodone prescriptions per county in PA, aside from Erie, “rural” counties, as defined by the 2010 U.S. Census after prescription rates were adjusted per 100,000 population, were actually the highest recipient of opioids by population (1).
Opioid Prescription Relative to Prescribers
As the top prescribers of opioid medications, primary care physicians (PCPs) contribute significantly to the prescriber role in the opioid epidemic. Guy et al. analyzed opioids dispensed from 90% of pharmacies across the U.S. from 2016-2017 using data from the IQVIA Prescriber Profile. This sample consisted of a total of 209.5 million opioid prescriptions from 970,902 prescribers across various specialties. Their analysis of the data showed that primary care physicians (family medicine, internal medicine, and general practice) accounted for 37.1% of opioid prescriptions, while non-physician providers (Physician Assistant and Nurse Practitioner) accounted for another 19.2% of prescriptions (5). Thus, these two groups comprised a majority percentage of opioid prescriptions dispensed during that year.
With this data, we hypothesized that counties in PA with the highest opioid prescription rates would have greater numbers of PCP and non-physician providers on average when compared to other PA counties. Data from the County Health Rankings and Roadmaps was used to further examine the relationship of population to provider ratio in the 16 counties that were in the top five for total or adjusted hydrocodone and oxycodone prescriptions rates in PA. Interestingly, we observed that half of these counties had fewer PCPs and non-physician providers than average to serve their ratio of the state population (Figure 5).
Figure 5: Ratio of Population to Primary Care Physician and Non-Physician Provider per County
Of the remaining eight counties, Allegheny and Blair had more PCP and non-physician providers than the PA average to serve their county population (4).
While PCPs are the highest prescribers of opioids based on specialty, this data suggests the imperative of access to care and a strong PCP relationship for patients managed with opioids. Additionally, for patients with opioid use disorder (OUD), PCPs can aid in monitoring, adjusting, and tapering dosages of opioid medications through more frequent follow-ups in contrast to prescribing larger or longer amounts (8). Access to a PCP also increases enrollment in medication-assisted treatment (MAT) for those with OUD in comparison to those without PCP access. Cole et al. examined Medicaid patients with OUD from 23 rural counties in PA and found that those with OUD and at least one PCP visit within the past year were more likely than those without a visit to receive MAT. Additionally, more buprenorphine and naltrexone prescriptions were filled. Furthermore, they found that rural Medicaid enrollees with OUD had substantial contact with their PCPs, averaging over 4 visits per person-year (3).
In summary, continued exploration of opioid prescription disparities across counties and specialties is crucial to identify trends and areas of improvement for providers. While urban areas of PA contribute objectively to a high percentage of opioid prescriptions, rural areas in the state also require important public health considerations. This is demonstrated in less-populated counties with decreased access to PCP and non-physician providers, yet high opioid prescription rates. PCPs are an integral part of opioid and OUD patient management. For example, with reduced access to PCPs who can counsel patients, frequently monitor opioid prescription dosages/amounts, and effectively taper dosages of opioids, patients may receive larger and longer prescriptions that place them at a higher risk of overdose or misuse and increase opioid circulation in the community. As the projected shortage of PCPs continues to increase, it is imperative that focus be placed on strategically training and placing PCPs to improve access to care and prevent patient overload in underserved areas (6). In addition, non-physician providers, such as nurse practitioners and physician assistants, are a key population for educating on evidenced-based opioid prescribing as well as managing patients on opioid medications or OUD treatment (5).
2014 Pulse of Pennsylvania Physician and Physician Assistant Workforce. https://www.health.pa.gov/topics/Documents/Programs/Workforce Reports/FINAL_2014 Pulse of Pennsylvania's Physician and Physician Assistant Workforce.pdf. Accessed June 18, 202.
U.S. State Opioid Dispensing Rates, 2019. Centers for Disease Control and Prevention. https://www.cdc.gov/drugoverdose/maps/rxstate2019.html. Accessed June 15, 2021.
Cole ES, Didomenico E, Cochran G, et al. The Role of Primary Care in Improving Access to Medication-Assisted Treatment for Rural Medicaid Enrollees with Opioid Use Disorder. Journal of General Internal Medicine. 2019;34(6):936-943. doi:10.1007/s11606-019-04943-6
County Health Rankings & Roadmaps. https://www.countyhealthrankings.org/app/pennsylvania/2021/measure/factors/131/data?sort=sc-0. Accessed June 18, 2021.